CSD Center for the Study of Democracy An Organized Research Unit University of California, Irvine www.democ.uci.edu

After more than a quarter of a century since the first third wave transition,1 we have seen the consolidation or likely consolidation of various democracies that resulted from different modes of transition. Does this mean that it is time to abandon the idea that the way in which a country transits from authoritarian rule matters? While we can perhaps discard mono-causal arguments that posit a necessary relationship between the mode of transition and democratic consolidation, this does not, however, preclude the possibility that there is a probabilistic relationship between the mode of transition and the likelihood of consolidation, that the challenges of consolidation are distinct, or that the mode of transition to democracy affects the type of democracy institutionalized. This article demonstrates that modes of transition indeed matter and that their effects merit further research. The literature on modes of transition has focused primarily on four tasks. First, many scholars have sought to cluster transitions into a limited set of modes, and to categorize recent and historic cases within these modes.2 Second, researchers have developed hypotheses about the effects of modes of transition on the likelihood of consolidation.3 Third, modes are often used to draw conclusions about the differing problems and tradeoffs to be faced during the transition and consolidation periods.4 Finally, some studies have begun to investigate the causal link between the mode of transition and the type of democracy that emerges.5 This study fits primarily into the latter two categories and focuses on the effects of pacted transitions – defined here as ones in which incumbent political elites from the authoritarian regime, who maintain control over the transition process, engage in multilateral negotiation and compromise with the democratic opposition. A review of the democratization literature reveals two general hypotheses about the effects of pacted transitions. On the one hand, theorists argue that pacted transitions offer the most viable path to the installation of democracy.6 On the other hand, many claim that pacted transitions may negatively affect democratic consolidation and deepening.7 Much of the literature on pact-making and pacted transitions claims that pacted transitions may depress access to positions of political leadership.8 For example, Przeworski argues that "the danger inherent in such substantive pacts is that they will become cartels of incumbents against contenders, cartels that restrict competition, bar access, and distribute the benefits of political power among the insiders. Democracy would then turn into a private project of leaders of some political parties and corporatist associations, an oligopoly in which leaders of some organizations collude to prevent outsiders from entering."9 Along these same lines, Karl argues that pacts may tend to demobilize new social forces and circumscribe the participation of certain actors in the future.10 However, little research has been conducted to determine whether this is indeed the case. Furthermore, if pacted transitions indeed produce higher levels of elite continuity, what is the causal mechanism? Do political pacts necessarily contain “agreements among leaders of political parties (or proto-parties) to…divide government offices among themselves independent of elections results,…and…exclude and, if need be, repress outsiders,” as Przeworski’s definition of political pacts implies?11 Finally, does the degree of elite continuity jeopardize the consolidation of democracy? Extremely limited access to positions of political leadership could clearly jeopardize democracy. However, elite continuity may be a positive sign of institutionalization12 and facilitate the development of a greater degree of expertise. Therefore, it is important not to draw conclusions prematurely. We need to discover if (and the degree to which) pacting depresses access to positions of political leadership, and, if so, determine the mechanism through which this occurs. We can then evaluate whether pacted transitions facilitate or hinder the consolidation of democracy. This article draws on a twelve-year cross-national study of the degree of elite continuity established after the Spanish (1976/77) pacted transition and the Argentine (1982/83) transition by collapse. Spain is the prototype of a pacted transition.13 If pact making produces the effect hypothesized, it should do so in the Spanish case. If the transition in Spain did not produce this effect, it would strongly refute the above hypothesis, despite being a single case. The comparison of Spain and Argentina varies the mode of transition. Upon his appointment by the Spanish monarch in July 1976, President Adolfo Suárez initiated a transition to democracy from within the regime and used the established rules of the Francoist state. The transition in Spain was negotiated by a reformist sector of the regime, led by Suárez, with hard-line sectors of the regime and with representatives of the democratic opposition. Neither the Suárez government nor the democratic opposition had the ability to entirely impose the terms of the transition. However, the Suárez government did not lose control of the process at any time. Parliamentary elections were held on June 15, 1977; the Unión de Centro Democrático (UCD), led by Suárez, emerged as the most successful party in the founding elections and Suárez formed a minority UCD government. The proximate cause and the formal beginning of the Argentine transition was the defeat of the Argentine military by the British in the Malvinas/Falkland Islands War (1982). However, the roots of the regime’s problems can be traced back to divisions within the military as early as 1980. While the regime was successful at penetrating society with fear, it was not as successful with regard to economic restructuring and growth. The transition in Argentina is best characterized as transition by collapse. In this study, transitions by collapse are ones in which the incumbent regime is left without the legitimacy to maintain authoritarian rule or to impose the terms of transition, and the outgoing authoritarian elites and the opposition do not engage in negotiation.14 The Argentine transition was not planned by the military regime, and the opposition did not force a transition to democracy. Rather, the transition resulted from the weakening of the regime due to their defeat in the war. The opposition did not push for the military’s immediate abandonment of power, partially due to the drastic economic situation facing the country. However, the military was left with no alternative but to begin preparing its withdrawal. The regime was unable to impose the terms of the transition, and the opposition successfully resisted the regime’s attempts to negotiate an amnesty for the acts committed in the dirty war. On October 30, 1983, Raúl Alfonsín was elected president of the nascent democracy. Under the new democracies in Spain and Argentina, the governing parties turned over government power to the principal opposition party within six years of the founding election. However, the empirical analysis of legislator (deputy) continuity in the lower houses reveals that

2 Spain (1977-1989/93) had significantly higher continuity than Argentina (1983-1993/5). Furthermore, the degree of legislator continuity, on average, has increased over time in Spain, and declined over time in Argentina.15 Is this related to the mode of transition? Did the higher level of continuity negatively impact democratic consolidation in Spain? This article makes three arguments. First, the higher level of deputy continuity in Spain vis-à-vis Argentina is, at least partially, the result of candidate selection procedures voluntarily adopted by the principal political parties. I will discuss three aspects of candidate selection: (1) rank-and-file v. elite selection, (2) regional v. national political party influence, and (3) the formal representation of internal party factions on electoral lists. The Spanish candidate selection procedures allowed a lower degree of rank-and-file participation and a higher degree of national political party elite control than the Argentine procedures. Furthermore, the Argentine parties permitted the formal representation of party factions on candidate lists whereas the Spanish parties did not. These procedures facilitated deputy continuity in Spain and encouraged renovation in Argentina. Second, the choice of candidate selection procedures in Spain was constrained by the pacted nature of the transition. Inter-party pacting required a reasonable expectation that the terms of the pact would be respected. Therefore, pacting required the existence of parties and party representatives willing to follow (or unable to oppose) the national party leadership involved in pacting. In short, successful pacting required disciplined political parties.16 This encouraged the adoption of internal party rules and regulations that facilitated pacting elites’ control over their parties and representatives. One vitally important means of ensuring party discipline is control over candidate selection. Therefore, the causal mechanisms that connect pacting and higher levels of legislator continuity are political party rules and regulations for candidate selection. This external constraint on the choice of candidate selection procedures did not exist in the Argentine case. Finally, I argue that rather than producing universally positive or negative effects, pacted transitions produce different challenges and hurdles than transitions by collapse. The pacted nature of the Spanish transition produced extremely hierarchical, rigid and disciplined political parties. It also produced greater levels of elite continuity. However, I also show that elite continuity did not have a net negative effect on democratic consolidation in Spain. I conclude that pacted transitions are not as prejudicial for democratic consolidation as initially feared, however there were birthmarks. On the other hand, Argentina experienced an extremely high level of elite renovation. While the Argentine political parties permit a greater degree of rank- and-file participation and a more powerful role for regional party organizations, Argentina suffers from a lack of institutionalized leadership in the Congress.

Deputy Continuity: Spain (1977-1989/93) and Argentina (1983-1993/95)

A comparison of the levels of continuity in the lower chambers is more interesting politically and more justifiable methodologically than a comparison of continuity in the upper chambers. First, the lower houses enjoy similar levels of authority, whereas the Argentine Senate is significantly more powerful than its Spanish counterpart.17 Second, the Spanish and Argentine electoral laws for the lower houses are very similar, whereas the election of senators occurs quite differently.18 Finally, the electoral laws for the lower houses have not fundamentally changed during the period

3 studied, whereas the laws regulating the election of senators have changed over time in the Spanish case.19 Both countries elect their lower houses on the basis of the d’Hondt method of proportional representation and both have a minimum threshold at the district level.20 There are no term limits in either case, and both countries adopted a system of closed-party lists. In Spain, the Congress of Deputies has 350 members directly elected from 50 multi-member districts and two single-member districts.21 The Spanish electoral laws also stipulate that each of the 50 provinces is guaranteed a minimum of two deputies.22 Deputies serve four-year terms (or until parliamentary elections are called), and all are renewed at the same time.23 Between 1983 and 1995, the Argentine Chamber of Deputies had 257 deputies (254 prior to 1991) directly elected from 24 multi-member districts. Similar to the Spanish case, the Argentine electoral laws provide for minimum levels of representation, namely that no district will have fewer than five deputies, and that no district will have fewer deputies than it had during the previous democratic period (1973-76). Deputies serve four-year terms, and half of the Chamber is renewed every two years.24 Table 1 presents reelection rates for Spanish (1977-89) and Argentine (1983-1993) deputies and the corresponding electoral volatility indexes.25 On average, 54 percent of Spanish deputies were reelected versus only 22 percent of the Argentine deputies. Therefore, Spain’s deputy continuity rate is thirty-two percentage points higher than Argentina’s.26

Table 1: Deputy Continuity and Electoral Volatility, Spain (1977-89) and Argentina (1983-93)

Spain Argentina % Reelected Volatility % Reelected Volatility 1977 - - 1983 - - 1979 55 7 1985 30 9 1982 40 43 1987 24 12 1986 57 12 1989 21 13 1989 62 6 1991 18 14 1993 15 8 Average 54 17 Average 22 11

Sources: Mercedes Alda Fernández and Lourdes López Nieto, “El Parlamento Español: 1977-1993” Revista de Estudios Políticos (Nueva Epoca) 81(July-September 1993), pp. 241-63; Eva Anduiza and Mónica Méndez, “Elecciones y Comportamiento Electoral,” in Manuel Alcántara and Antonia Martínez, eds., Política y Gobierno en España (Valencia: Tirant lo Blanch, 1997). The Argentine reelection rates were calculated by the author based on deputy lists provided by the Argentine Chamber of Deputies. Kenneth M. Roberts and Erik Wibbels, University of New Mexico, provided the Argentine volatility indexes.

The higher continuity rate in the Spanish Congress of Deputies holds true regardless of which election we examine. For example, Spain experienced the lowest degree of deputy continuity in 1982 (40 percent) when the entire party system realigned. This election produced the dramatic electoral defeat of the governing UCD. Electoral support for the UCD declined more than 28 percentage points. This decline resulted not only in the turn over of government to the opposition Partido Socialista Obrero Español (PSOE), but also in the rise of the Alianza Popular (AP) as the largest opposition party. If we compare this to the highest Argentine deputy

4 continuity rate in 1985 (30 percent), we still find that the Spanish continuity rate is 10 points higher.27 Electoral volatility cannot account for the cross-national difference. A comparison of the formal levels of electoral volatility, using the Pedersen index, confirms the higher level of electoral volatility in Spain.28 The average volatility index for elections to the Spanish Congress of Deputies during this period is seventeen compared to eleven for Argentine Chamber.29 The six- point difference clearly demonstrates the higher level of electoral volatility in the Spanish case. If we were to predict the level of continuity solely on the basis of electoral volatility, we would predict that Spain would have a lower degree of continuity than Argentina.30 Furthermore, as can be seen in Figure 1, aside from a decline of continuity in 1982, reelection rates climbed in the Spanish case– reaching 64 percent in 1993.31 In contrast, reelection rates consistently fell in the Argentine case – reaching a low of 15 percent in 1993.

Figure 1: Comparison of Deputy Re-election Rates in Argentina and Spain

70%

60%

50%

40%

30%

20%

10%

0% Election 2 Election 3 Election 4 Election 5 Election 6

Spa Deputy Arg Deputy

Election 2 Election 3 Election 4 Election 5 Election 6 Spanish Deputy 55% 40% 57% 62% 64% Argentine Deputy 30% 24% 21% 18% 15%

This pattern remains when we take out the consecutive nature of the reelection measure and determine the number of terms that each deputy served. As can be seen in Table 2, 558 or 36 percent of the Spanish deputies served only one term, while 70 percent of the Argentine deputies did so.32 Furthermore, forty-six Spanish deputies held their seat throughout the entire period whereas this number drops to just four of the Argentine deputies.33

5 Table 2: Terms in Office, Deputies, Spain (1977-1989) and Argentina (1983-1995)

1 Term 2 Terms 3 Terms 4 Terms Entire Period N Spain 558 277 81 46 46 1539 Argentina 706 119 21 1 4 1011

Sources: Elaborated by the author based on deputy lists provided by the Spanish, Congress of Deputies and the Argentine Chamber of Deputies.

What accounts for the cross-national variation in levels of deputy continuity? The next section argues that the answer, at least partially, lies with the candidate selection procedures.

Candidate Selection Procedures and Deputy Continuity

The electoral laws in both countries stipulate the use of closed-party lists for elections to the lower house, which means that the political parties control which candidates appear on the election ballot and the order in which they appear. The voter cannot designate whether her vote goes for a particular candidate, alter the order of the list or cross out any candidates. The voter must vote for the entire list as presented by the party. Closed-list systems give parties the greatest degree of control over candidates and, in turn, encourage party discipline.34 For our purposes, closed-lists also allow parties a great deal of influence over the level of candidate and deputy continuity. Despite the fact that the electoral laws in both countries stipulate the use of closed party- lists for the election of deputies, this does not determine the means by which the parties will choose those candidates. According to Richard Katz,

Candidate selection is a vital activity in the life of any political party. It is the primary screening device in the process through which the party in public office is reproduced. As such, it raises central questions about the ideological and sociological identities of the party as a whole. Moreover, because different modes of selection are likely to privilege different elements of the party and different types of candidates, they may raise questions about the nature of the party as an organization as well.35

I argue that the distinct levels of continuity are partially the result of internal party rules and regulations that were adopted by the political parties. During the period studied, both countries produced party systems in which two parties dominated. In the 1977 and 1979 Spanish parliamentary elections, the UCD and the PSOE captured an average of 82 percent of the seats in Congress. Following the 1982 elections, the two largest political parties changed with the AP replacing the UCD. In the 1982 and 1986 elections, they captured 85 percent of the seats.36 The discussion of the Spanish parties will largely focus on the PSOE and UCD as these parties emerged out of the transition as the principal political parties and were the main actors involved in transition pacting. In Argentina, the two largest parties, the Partido Justicialista (PJ) and the Unión Cívica Radical (UCR), attained a combined maximum of 95 percent of the seats in the Chamber of Deputies and a minimum of 80 percent.37

6 I will discuss three aspects of candidate selection: (1) rank-and-file v. elite selection, (2) regional v. national political party influence, and (3) the formal representation of internal party factions on electoral lists. The Spanish candidate selection procedures allow a lower degree of rank-and-file participation and a higher degree of national political party elite control than the Argentine procedures. Furthermore, the Argentine parties permit the formal representation of party factions on candidate lists whereas the Spanish party rules do not. These procedures facilitate deputy continuity in Spain and encourage renovation in Argentina.38 In Argentina, the provincial-level party organizations generally have a greater degree of influence over the selection of party candidates than the national party.39 The influence of the provincial-level parties is enhanced when the party’s leader is not the Argentine president and when the local party leadership is united.40 The local party boss plays a particularly important role in candidate selection and contributes to the high rate of deputy turnover. According to Jones, et al, “the threat of challenge by popular legislators provides local party bosses with a strong incentive to reduce the national and provincial visibility of their local underlings by rotating them among the various jobs the provincial party can offer.”41 However, the party leadership does not always determine candidate selection. The PJ and the UCR also permit direct participation by rank-and-file party members through the use of party primaries to determine candidate lists. Party primaries may be used when the party cannot agree on a common list.42 Furthermore, sitting UCR deputies, competing in internal primaries, must attain 2/3 of the vote in order to be, again, presented on the party list.43 While the UCR has a history of using primary elections to choose candidates, primaries are a departure for the PJ - the party rarely used primaries to select candidates before 1985.44 It is thus not a mere resurrection of previous procedures. The use of primaries came about in response to internal challengers, the renovadores, to the then existing party leadership. The calls for internal democracy gained support from many party members and affiliates particularly after the PJ’s historic loss to the UCR in the 1983 presidential elections.45 According to De Luca, Jones and Tula, the use of primaries progressively increased between 1983 and 1993.46 In 1983, 13 percent of UCR provincial party branches used primaries to choose their candidates. A majority of UCR provincial party branches used this method by 1985, and by 1993, this had increased to 71 percent. In 1983, no PJ provincial party branch used primaries to choose their candidates. However, a majority of provinces used this method by 1989, and 71 percent used primaries in 1993. On average, 67 percent of UCR provincial party branches used primaries between 1985 and 1993. The PJ average is 50 percent. The use of open versus closed primaries has also been on the rise in both parties.47 Finally, the Argentine parties often incorporate the representation of minority party factions for both party offices and candidacies for public office.48 There are a variety of formulas used to allocate list positions after a primary. According to De Luca, Jones and Tula,

These include winner-take-all arrangements in which the list that wins the plurality of the vote in the primary wins all of the positions and semiproportional methods in which the runner-up list (if it surpasses a threshold, normally 25% of the vote) receives every fourth (PJ in large districts) or third (UCR of PJ in small districts) position on the party list. Other district-level parties use the d’Hondt PR formula, often with a 25% threshold.49

7 Furthermore, when primaries to determine deputy lists occur, the minority list usually receives enough votes to obtain a position on the list.50 In contrast, the principal Spanish political parties, the UCD and PSOE, did not establish a system of primaries to choose their candidates, and there were no formal hurdles placed on the reelection of sitting deputies.51 There has also been a tendency to formalize the national leadership’s control of candidate selection vis-à-vis the provincial leadership.52 Finally, the parties established no formal means of representing minority factions.53 Though PSOE party statutes did not formally permit the national level party organization to alter candidate lists until following the 1979 elections, the national party leadership in 1979 intervened to “influence the composition of party electoral lists, which were supposedly the responsibility of provincial federations.”54 In 1979, the party Congress revised the party statutes to formally allow party elites at the national level to alter candidate lists. According to party statutes, the party hierarchies at both the local and regional levels develop preliminary candidate lists. These lists require the approval of the national level party organization, the Federal Committee.55 If there is a disagreement between the national and regional party organizations, the final decision is left to the Federal Committee. Though the Federal Committee is formally the most important party body after the party Congress, “in practice, its role has been reduced to discussing and ratifying Executive Committee postures and decisions.”56 These candidate selection procedures remained in place for the elections between 1979 and 1996. Also, national elite control of candidate selection is a historical departure for the PSOE. During the Second Spanish Republic (1936-39), PSOE statutes stipulated that an assembly of affiliates within a district would meet to designate candidates for national deputy.57 The UCD is a very interesting case. Founded during the transition as an electoral coalition of various formally recognized political parties, it functioned differently from the PSOE. The candidate selection process, at least informally, required the inclusion of candidates from the distinct political families that made up the UCD. Nevertheless, the leadership of Adolfo Suárez, the last president of the authoritarian regime and leader of the UCD, dominated this process of candidate selection.58 According to Hopkin, “ultimately the Suárez government made all of the key decisions on the electoral lists,” and he quotes Calvo-Sotelo – whom Suárez designated to negotiate the 1977 lists - to confirm this:

It was so impossible to reach agreements in the first two or three days that I decided to take a dictatorial line, because the deadline for the nomination of candidates was approaching…I proposed that one person, which would be me, should decide the lists in all 52 provinces. At that moment Suárez had all the votes, and I represented him, so they had no choice but to accept those conditions.59

This control over the candidate lists led to significant revisions of candidates initially proposed by the nascent coalition, Centro Democrático.60 Therefore, for the first elections, Suárez’s control of candidate selection was largely informal. Immediately following the founding election, the UCD took formal steps to transform the coalition into a unified political party, and held its first conference in 1978. The party statutes passed at the conference formalized national political control of candidate selection. The statutes indicated that it was the responsibility of the Provincial Executive Committees “to propose…the names of the candidates for Deputy and Senator for their province.” However, the party statutes

8 also gave the party president formal control over “electoral matters.”61 Hopkin states that this would allow the party leader “to exercise control over candidates presented in elections, and therefore in practice choose (voters willing) the elected representatives of the party.”62 Suárez’s control was not merely formal. According to an interview conducted by Hopkin, for the 1979 lists “there was an electoral committee whose views were heard, but in the end Suárez himself and the secretary general decided – that was a matter which Suárez paid a lot of attention to.”63 Furthermore, provincial participation in the process was largely ignored.64 The party statutes passed at the 1981 party conference changed the candidate selection procedures slightly. Article 51.2.C. re-stated that the Provincial Executive Committees “propose…the names of the candidates for Deputy and Senator.” However, Article 22.3 stated that “a National Electoral Committee will be formed to decide on electoral matters…and will be made up of the President and Secretary General of the Party, who will be its President and Secretary, respectively, and three members elected by the Executive Committee of the Party from amongst its members.”65 Whereas the party president had previously controlled electoral matters, they were now to be decided by a National Electoral Committee, largely under his control. Similar to the PSOE, the control over candidate selection was concentrated in the hands of the national party leadership.66 Finally, minority party factions were not formally guaranteed positions on PSOE and UCD candidate lists. This provides the existing party leadership with an additional degree of insulation from internal challengers. Though internal party dissidents in the PSOE, the críticos, attempted to introduce the representation of minority factions at the party congress in 1979, they were unable to do so.67 The PSOE statutes explicitly stated that the party would not tolerate ‘organized tendencies,’68 which Gillespie defines as “factions with independent structures and the ambition to take control of the party.”69 This was the case throughout the period studied.70 Despite the fact that the UCD had begun as an electoral coalition of independent political parties, Article 12.3 of the party statutes forbid the existence of organized groups or factions.71 Though factions existed de facto, along the lines of the founding parties, the inclusion of various groups on party lists and in governmental positions was left to informal negotiation amongst party barons.72 The fact that ultimate control over candidate lists resided with the national party elites and that minority factions were not guaranteed representation provided a degree of insulation and control that was not available to the Argentine national party leaders. In Argentina, the party rules provided more room for internal challengers to affect outcomes and a greater degree of provincial party influence. Though list systems may facilitate party discipline, the rules adopted by the Spanish parties meant that the national party elites had a greater degree of control over the direction of party strategies and positions. In order to demonstrate that these candidate selection procedures indeed affected the degree of continuity, I measure the degree of candidate continuity.73 I conducted the analysis by political party and at each point in time the measures include the two main political parties.74 Given the change in the dominant Spanish political parties, the measures for 1979 and 1982 include continuity figures for the UCD and PSOE, and the 1986 and 1989 measures include continuity figures for the PSOE and AP/PP. Table 3 presents the average proportion of candidates re-nominated by the two largest parties in each country.75 The data show that the principal Spanish parties re-nominate significantly more candidates than the principal Argentine parties, 30 percent versus 18 percent.76

9 Table 3: Candidate Continuity, Two Largest Parties, Congress of Deputies, Spain (1979-89) and Chamber of Deputies, Argentina (1985-93)

Spain Argentina % Re-nominated % Re-nominated 1979 32.5 1985 24.5 1982 25.5 1987 19.5 1986 31.5 1989 16 1989 31.5 1991 13 1993 16 Average 30.0 Average 17.8

Source: Calculated by the author based on candidate lists published in the Spanish Boletín Oficial del Estado and collected from the Departamento de Estadísticas, Dirección Nacional Electoral, Ministerio del Interior, Argentina.

Moreover, continuity in Spain is under-stated in these results. This measure does not take into account candidates that left one of the main parties to become a candidate for another political party. For example, the PSOE, between 1979 and 1986, incorporated seven UCD candidates. AP, during the same period, nominated forty-one candidates that had previously appeared on UCD lists. Though candidates in Argentina may defect from one political party to join another, they are unlikely to switch from the PJ to the UCR or vice versa, given the strong personal and historical connection that many of these deputies have to their party.77 Therefore, the gap between the two cases is probably larger than the above results indicate. We find further evidence of the importance of candidate selection procedures by looking at the Argentine data over time. Though this measure does not control for the influence of other factors, the relationship between party primaries and candidate renovation in Argentina is illustrated by the fact the use of primaries has increased over time, while candidate continuity has largely declined.

Table 4: Candidate Continuity by Party, Argentina (1985-1993)

Averag 1985 1987 1989 1991 1993 e UCR 32% 23% 19% 11% 13% 19.6% PJ 17% 16% 13% 15% 19% 16.0% Average 24.5% 19.5% 16.0% 13.0% 16.0%

Source: Calculated by the author based on candidate lists collected from the Departamento de Estadísticas, Dirección Nacional Electoral, Ministerio del Interior, Argentina.

Unlike the Argentine parties that progressively increased the voice of rank-and-file members through the use of primaries, the Spanish parties made no attempt to do so. Rather, the parties progressively reduced the participation of regional party organizations (which are most likely to be influenced by rank-and-file members) by centralizing the selection of party candidates in the hands of elites at the national level.

10 Pacted Transitions and Political Party Rules

The argument that party rules for the selection of candidates affect the degree of continuity may not be particularly surprising or counter-intuitive. However, why parties in new democracies choose to adopt particular candidate selection procedures is an interesting and unexplored question. Historic parties do not always maintain previous procedures. Furthermore, new parties develop without a historic tradition. Why are certain procedures chosen over others? Why are they maintained or altered? I argue that the pacted nature of the Spanish transition to democracy constrained the choice of candidate selection procedures; the procedures adopted permitted a greater degree of control by national party elites. While some scholars have explored the effects of the mode of transition on political parties in Spain,78 to my knowledge, no one has directly addressed the effect of the mode of transition on the choice of candidate selection procedures, and how those procedures, in turn, facilitate elite continuity. Successful pacting required that the terms of the pact be respected. As Stepan argues, “party pacts by their very nature have two indispensable requirements: first, leaders with the organizational and ideological capacity to negotiate a grand coalition among themselves; second, the allegiance of their political followers to the terms of the pact.”79 Pacting, by definition, also required compromise on both sides. Compromise required that the main negotiating partners reduce the distance between their positions. This resulted in the move to the right by the PSOE and the corresponding move to the left by the UCD.80 This moderation did not enjoy universal support. Pacting elites needed to count on party executive committees, governmental appointees and parliamentary groups that supported (or could not oppose) their actions. One implication is that pacting encouraged the adoption of internal party rules and regulations that facilitated pacting elites’ control over their parties. Pacting elites had to organize their parties in such a way as to isolate internal dissent. The literature on Spanish political parties illustrates that this was indeed the case. According to Capo, pacting allowed the participants to think in terms of adversaries instead of enemies, which was positive for democratic consolidation.81 However, he also points out that the form of reaching agreements, pacted by the party leadership, strengthened internal party discipline. The parties adopted internal party rules and parliamentary procedures to foster party discipline.82 Capo also argues that pacting cut the ties between elites, party activists and the electorate, and facilitated a distancing between the state and society.83 The continuation of elite negotiations during the 1977-79 period meant that leaders put little emphasis on mobilization and that party organizational development occurred from the top down. Caciagli concurs that the mode of transition “conditioned the parties’ capacity for solid political and organisational development.”84 This study is consistent with both of these arguments. The candidate selection procedures adopted in Spain were those most likely to produce party discipline. Furthermore, these procedures did not encourage mobilization. Control over the party in parliament was fundamental to the success of pacting, and national elite control over candidate selection meant that their supporters were more likely to maintain (or attain) positions in parliament. Though from distinct starting points, the political parties involved in transition pacting rapidly implemented mechanisms to encourage party discipline. These mechanisms also facilitated elite continuity. These candidate selection procedures were a fundamental part of the hierarchicalization of Spanish political parties. Prior to tracing the influence of the mode of transition on political

11 parties in Spain, it is important to recall the historical context in which party development took place. The formal beginning of the transition started with the appointment of Adolfo Suárez by King Juan Carlos on July 3, 1976. The transition can be broken down into two phases. The first phase began with Suárez’s appointment and ended with the founding elections on June 15, 1977. Upon his appointment, Suárez initiated a transition to democracy from within the regime and used the established rules of the Francoist state. The initial days of this phase of the transition were marked by negotiations with various elites within the regime. True negotiations with the opposition began following the approval of the Law for Political Reform, the government’s framework for political transition. The second phase began following the 1977 founding elections and ended with the second parliamentary elections in March 1979. Suárez’s UCD emerged as the largest party and formed a minority government. The PSOE was the largest party of the left and the principal opposition party. This second phase of the transition coincided with the consensual negotiations of socio- economic accords (Pactos de la Moncloa), regional devolution of power to the historic regions of Catalunya (Cataluña) and Euskadi (País Vasco), and the new constitution approved in 1978. This period was marked by the moderate stances of the principal political parties. Following the approval of the Constitution by referendum in December 1978, the government called new elections for March 1979. The March 1979 elections largely confirmed the results of the founding elections, and the UCD once again formed a minority government between 1979 and 1982. This period corresponded with the ‘crisis of the parties’ – during which the PCE, PSOE and UCD suffered severe internal divisions. The period also saw the resignation of Adolfo Suárez and the dramatic coup attempt on February 23, 1981, when members of the civil guard captured the Congress of Deputies during the investiture vote of prime ministerial candidate, Leopoldo Calvo-Sotelo. The coup was thwarted, and Calvo-Sotelo was elected Prime Minister. In 1982, elections were called, which led to the victory of the PSOE with an absolute majority in the parliament. The establishment of elite control over candidate selection in the PSOE reflects a larger process of the concentration of power in the hands of pacting elites. Unlike the UCD, the PSOE was not a new political party. In fact, the PSOE was the largest party of the left in the Spanish Second Republic (1931-6). However, organizationally, the party was tremendously weak. It claimed two thousand members in 1974, which rose to only eight thousand in 1976.85 Furthermore, the party leadership had been in exile in France, and dominated by exiled leaders. This changed in 1974 when Felipe González was elected Secretary General, definitively taking control of the PSOE from the exiled leadership. The new executive formed under his leadership was made up exclusively of representatives of the party in Spain.86 By 1976, control of the party organization was largely concentrated in the hands of the Sevillanos, led by González and Alfonso Guerra,87 who consolidated the party apparatus and discipline in the interior.88 After its legalization in 1977, González and Guerra gained increasing influence over the Executive Committee, which in turn dominated the much larger and statutorily more important Federal Committee.89 Though the parliamentary group elected in 1977 was heterogeneous, for the most part it was inclined toward pragmatism.90 During the 1977/79 negotiations with the UCD government, the PSOE took considerable steps to assure the maintenance of the party line established by pacting elites. Felipe González led the Socialist Parliamentary group as spokesperson in the Congress of Deputies, and González made decisions largely without consulting the party or

12 parliamentary group. This included the signing of the Pactos de la Moncloa and the compromises reached with the UCD on the Constitution.91 This period was also marked by expulsions from the party of more radical members and the incorporation of new, less politically experienced members. According to Gillespie,

Altogether, the left suffered several hundred expulsions and suspensions. Other militants left in disgust, unable to stomach the moderate course being steered by the party leaders. Disciplinary action against dissenters, often for criticizing leading members, making declarations to the media or selling unofficial socialist papers, was taken with no prevarication and was often initiated by the central party apparatus headed by Alfonso Guerra.92

The party executive also intervened in the selection of candidate lists for the Cortes in 1979, which generated a great degree of controversy within the party.93 Though the formal transition ended with the approval of the 1978 Constitution and the calling of elections under the new rules of the game, the transition took place without dismantling or reforming the coercive and military apparatus of the state. This meant that the military and security forces were still significant actors - powerfully illustrated by the attempted coup in February 1981. Therefore, for the parties of the left, the move to the right and the importance of party discipline were not simply due to reaching agreements with the UCD. They also had to be concerned with the military’s reactions to the process of democratization and to a possible future PSOE government. This partially explains why the PSOE continued the process of moderation and increasing hierarchicalization of the party following the formal end of the transition in 1979. Significant changes in party organization occurred between 1979 and 1982. First, the party changed the manner in which it elected delegates to the Federal Congress. A majority list system, in which the most voted list sent all of the delegates to the Federal Congress, was used until 1984. However, prior to 1979, local party branches directly elected the delegates. After the May Congress, the local branches elected delegates to the regional Congresses, which in turn elected the delegates to the Federal Congress. The establishment of indirect elections meant that the central party apparatus was better able to influence the choice of congressional delegates.94 This influence was significant because the Federal Congress elects the executive bodies of the party, and approves party statutes and resolutions. Furthermore, the 1979 statutes allowed the regional Congresses to determine whether their delegates would be required to vote as a block. In 1979, the party Congress also revised the statutes to formally allow party elites at the national level to alter candidate lists. These candidate selection procedures remained in place for the elections between 1979 and 1996. Despite internal controversy over these issues, the party also maintained its prohibition of organized internal factions and a majority list electoral system for Executive Committee elections. Finally, in 1981, power was further concentrated in a Permanent Committee within the Executive Committee. The Permanent Committee included the Secretary General, Vice-secretary General, and the Area Secretaries.95 González and Guerra dominated the Permanent Committee, which was responsible for the daily functioning of the party.96 As the discussion of PSOE party development illustrates, the party rules permitted a great degree of leadership control over the party and its representatives during the very delicate period of transition and the first years of democracy. Candidate selection was one of the key mechanisms used to bolster the position of the party leadership, and, in turn, fostered continuity.

13 The effect of transition pacting on party rules is also demonstrated by UCD party development. The development of the UCD coincided with the first phase of the transition, from Suárez’s appointment by King Juan Carlos on July 3, 1976 to the founding elections of June 15, 1977. The UCD began as the Centro Democrático, which was a nascent electoral coalition of center-right proto-parties largely organized by the Partido Popular. The Partido Popular was presented to the public in December 1976 and was led by former regime ministers Pío Cabanillas and José María de Areilza.97 The parties of the Centro Democrático organized in response to regime reforms and around reformist Francoist politicians and members of the moderate-right opposition to the Franco regime. However, these were parties in name only - none of them had a mass following. On May 3, 1977 and following the successful initiation of regime reform from within, Suárez publicly announced that he would run as the prime ministerial candidate for the Centro Democrático, which was subsequently changed to Unión de Centro Democrático. Suárez’s decision to join the Centro Democrático had several organizational consequences. Suárez and his followers were given a great degree of control over the coalition. He was able to demand that his potential challenger, José María Areilza, be removed from the coalition’s leadership. In April 1977, Leopoldo Calvo-Sotelo, Suárez’s Minister of Public Works, resigned his position in order to organize the UCD. Calvo-Sotelo approved the UCD’s foundational text and gained a great degree of control over the selection of candidates for the upcoming parliamentary elections.98 According to Jonathan Hopkin, Suárez dominated “the process of candidate selection, imposing candidates close to him and other key members of the government (especially Martín Villa). This left many of the centre leaders unhappy at their representation in parliament, and limited their ability to influence the development of the UCD.”99 Three of the original fifteen parties left the coalition in protest.100 By influencing the makeup of candidate lists, Suárez attempted to assure his domination of the coalition and the incorporation of his followers. However, given the UCD’s coalitional nature, the various political families gained representation on the lists and subsequently in the Congress of Deputies. The second phase of the transition between June 15, 1977 and March 1979 was marked by the transformation of the coalition into a formal political party with a hierarchical structure. The organizational model of the UCD at this time was characterized by the predominance of the government, and more precisely by Suárez and a small group of supporters.101 Immediately following the founding elections, the various parties committed to turning the UCD into a formal political party, and the deputies and senators formed a single parliamentary group in each chamber.102 The coalition partners also “committed themselves to vote as a group on the defense of democracy, the monarchy, the unity of the state, the recognition of the autonomies and the social market economy.”103 In order to facilitate integration, Suárez gave ministerial positions to the leaders of the various coalition groups in July 1977. The first party Congress in October 1978 formalized the centralization of power. The party adopted a presidential structure, and confirmed Suárez as the uncontested leader of the party. The party statutes established a majoritarian electoral system for the party Executive Committee, on which Suárez was able to place personal friends.104 Also, government ministers dominated the Executive Committee. According to Hopkin, this meant “the executive committee was in effect the Council of Ministers in another guise…This allowed Suárez to share decision- making with ministers whom he himself had chosen, rather than with an executive which was

14 more representative of the original UCD parties.”105 The statutes also established the predominance of the national party organization over the provincial and regional organizations.106 Party factions were prohibited, and Suárez demanded and received a greater degree of control over the candidate lists for the 1979 elections. Once again, a group close to Adolfo Suárez determined the candidate lists for the 1979 elections. In many cases, the leaders of the original parties were placed in locations on the lists that did not meet their expectations.107 Following the electoral victory of 1979, Suárez did not follow the previous pattern of incorporating the leaders of the original coalition parties into the government, but rather chose personal supporters.108 Why did this rapid transformation occur? Also, why did the party choose this organizational model as opposed to one that accommodated the diverse nature of its origins? In order to explain the UCD’s decision to form a single parliamentary group and to vote together, Esteban and López Guerra argue that at a time when the dictatorship was being dismantled and the constitution was being written, “the existence of UCD as a unified organization was the only alternative facing the disciplined parties of the left: its division would have meant leaving the moderate Spanish right without a qualified spokesperson, which would have possibly affected the entire process of writing the Constitution.”109 Along these same lines, the requirements of transition pacting facilitated the rapid hierarchicalization of the PSOE and the transformation of the UCD from a diverse coalition into a hierarchically organized political party. A fundamental aspect of that transformation was national party elite control over candidate selection. Can the length of the previous regime better explain the candidate selection procedures adopted by the Spanish parties? After more than forty years of authoritarian rule, the Spanish political parties were poorly organized and not deeply rooted in Spanish society. In fact, Caciagli states that “organised parties in the real sense did not exist in the country, except for the communists.”110 Furthermore, the Spanish political parties were legalized just a few months prior to the first election. Without the existence of a significant number of party members or infrastructure, the parties may have been less likely to adopt a primary system or a system in which regional party branches played a larger role. The situation may have encouraged a greater degree of national elite control. This can also be partially applied to the Argentine case. Though Argentina had a more recent experience with democracy, the Argentine parties also faced the challenges of party re-organization and the development of party infrastructure. This is illustrated by the fact that only 13 percent of UCR provincial party branches used primaries to select their candidates for deputy in the first election in 1983,111 and this is a party with a history of primaries. However, the length of the previous regime in Spain cannot explain why the candidate selection procedures were maintained, or why national elite control was formalized. The fact that Spanish parties were not deeply rooted in Spanish society would not preclude the use of party primaries or a system of candidate selection that concentrated power at the regional level in future elections. Furthermore, it could certainly be argued that the opportunity for participation within the parties could strengthen party organization, membership development and electoral prospects. Though political parties in Spain were very weak at the time of transition, distinct strategies were compatible with their organizational development. Caciagli states that “the major responsibility of the Spanish parties vis-à-vis society during this period was to regulate the mode of behavior of their elites,” which he calls “the style of politics.”112 This article argues that a fundamental means of regulating elite behavior was controlling who would represent the parties. The PSOE and UCD party leadership, led by Felipe

15 González and Adolfo Suárez, respectively, had already begun pacting prior to the first democratic elections. However, following the 1977 elections there were fundamental issues to be addressed, such as the writing of a new constitution, demands for regional autonomy, a growing economic crisis, military reform and rising terrorism. In order to continue pacting and avoid a backlash from those opposed to the transition (and from those within their own parties that disagreed with the consensual strategy), they needed to maintain control of their parties. They had to ensure that their supporters were represented in their parliamentary groups, which needed to ratify many of the agreements. If they did not have support from their parliamentary groups, the transition process would be put in jeopardy. The fact that pacting elites were concerned with maintaining parliamentary groups that would support their strategies and positions combined with the degree of control over candidate selection explains the relationship between pacting and the higher degree of elite continuity in Spain vis-à-vis Argentina. Pacting explains not only why elites (versus the rank-and-file) dominated the candidate selection process, but also why candidate selection was ultimately decided by national political elites – these were the very individuals responsible for negotiating the transition.

Transitions, Levels of Elite Continuity and Democratic Consolidation

This article demonstrates that the pacted transition in Spain produced a higher level of elite continuity than the Argentine transition by collapse. I ague that transition pact making provided an external constraint on the choice of candidate selection procedures in Spain. Pacting encouraged the adoption of internal party rules and regulations that facilitated pacting elites’ control over their parties and representatives. One powerful means of ensuring party discipline was control over candidate selection. The stakes involved with maintaining party discipline encouraged national party-elite control of candidate selection, the prohibition of internal factions, and the limitation of rank-and-file participation. These institutional rules facilitated elite continuity. The constraints of transition pact making, on the other hand, were not present in the Argentine case. Rather, internal party development was left largely to internal party dynamics and party competition. Perhaps due to the lack of a powerful external constraint, the Argentine parties were more likely to adopt institutions that permitted a greater degree of internal competition to determine the direction of party politics and party leadership. The rules adopted, in turn, encouraged elite renovation. Did the pacted nature of the Spanish transition and the higher level of elite continuity negatively affect democratic consolidation? There are several reasons to conclude that pacted transitions are not as prejudicial for democratic consolidation as initially feared. First, the absolute level of deputy continuity in Spain (54% reelection) is not high compared to established democracies such as the United States. Second, contrary to Przeworski’s definition, the political pact in Spain did not include an agreement “among leaders of political parties (or proto-parties) to…divide government offices among themselves independent of elections results,…and…exclude and, if need be, repress outsiders.”113 In other words, outcomes were not guaranteed. Third, the higher level of overall elite continuity in the Spanish case did not preclude renovation vis-à-vis the previous authoritarian regime, nor did it uniformly impede the renovation of the pacting leadership and the transfer of governing power to the opposition left. In the 1982

16 parliamentary elections, the then governing UCD attained only 11 seats (3%) in the Congress of Deputies. Suárez’s new electoral vehicle, the Centro Democrático y Social, only attained two seats, and could not form its own parliamentary group. The 1982 elections also marked the first turnover of political power from the governing UCD to the opposition PSOE. Governing power was placed in the hands of a political party unassociated with the previous authoritarian regime. Finally, elite continuity facilitated elite negotiations and compromise on key political issues.114 Pact making began under the previous authoritarian regime when negotiations occurred between the Suárez government and the leadership of the democratic opposition parties. The negotiations led to the progressive dismantling of the Francoist institutions and democratic elections in 1977. Suárez led the UCD to a victory in the founding elections of 1977. The same individuals involved in pre-election pact making became prominent political figures in the post- election period. The success of previous negotiations carried over into the post-election period. These leaders proceeded to sign socio-economic accords, the Moncloa Pacts, design a consensus- based constitution, set up provisional governments in the historic regions of Spain, and collaborate on the passage of legislation in the parliament. The continuity of political elite actors facilitated the maintenance of this political style and the resolution of crucial political issues during the initial years of fragile democracy. However, there were also related negative effects. First, higher levels of continuity under the new democracy went along with the hierarchicalization of the Spanish political parties. While this did not prevent the consolidation of democracy, it limited the impact of rank-and-file members on party politics and likely increased the distance between the public and the political parties. The mode of transition, therefore, had a powerful affect on the type of democracy institutionalized. The Argentine transition, on the other hand, produced distinct challenges for democratic consolidation. Argentina experienced an extremely high level of elite renovation. While the Argentine political parties permit a greater degree of rank-and-file participation and a more powerful role for regional party organizations, Argentina suffers from a lack of institutionalized leadership in the Congress, which may limit the possibility of cross-party collaboration and exacerbates Congress’s weakness vis-à-vis the president. This article has identified an important causal variable, the mode of transition, which affects the adoption of internal party rules and procedures. Future research needs to explore the effects of additional modes of transition on internal party politics. Future research must also identify additional causal variables that explain why political parties in transitional settings adopt particular internal party rules and procedures. Finally, we to further explore how internal party organization affects democratic consolidation and deepening. Very little research has been conducted on these fascinating and important subjects, particularly in new democracies.

17 Endnotes

1 Samuel P. Huntington, The Third Wave (Norman: University of Oklahoma Press, 1991). 2 Robert A. Dahl, Polyarchy: Participation and Opposition (New Haven: Yale University Press, 1971); Juan J. Linz, “Crisis Breakdown, and Reequilibration,” in Juan J. Linz and Alfred Stepan, eds., The Breakdown of Democratic Regimes (Baltimore and London: The Johns Hopkins University Press, 1978); Juan J. Linz, “Comment: Some Comparative Thoughts on the Transition to Democracy in Portugal and Spain,” in Jorge Braga de Macedo and Simon Serfaty, eds., Portugal Since the Revolution: Economic and Political Perspectives (Boulder, Colorado: Westview Press, 1981); Donald Share, Transition Through Transaction: The Politics of Democratization in Spain, 1975-1977 (Doctoral Dissertation, Stanford University, 1984); Donald Share and Scott Mainwaring, “Transitions Through Transaction: Democratization in Brazil and Spain,” in Wayne A. Selcher, ed., Political Liberalization in Brazil (Boulder and London: Westview Press, 1986); Donald Share, “Transitions to Democracy and Transition Through Transaction,” Comparative Political Studies 19 (January 1987), pp. 525-48; Juan J. Linz, “Transitions to Democracy,” The Washington Quarterly (Summer 1990), pp. 143-64; Terry Lynn Karl, "Dilemmas of Democratization in Latin America," Comparative Politics 23 (October 1990), pp. 1-23; Huntington, The Third Wave; Terry Lynn Karl and Philippe C. Schmitter, "Modes of Transition in Southern and Eastern Europe, Southern and Central America," International Social Science Journal (May 1991), pp. 269-84; Gerardo L. Munck and Carol Skalnik Leff, “Modes of Transition and Democratization: South America and Eastern Europe in Comparative Perspective,” Comparative Politics 29 (April 1997), pp. 343-62. 3 Dahl, pp. 41-45; Samuel P. Huntington, “Will More Countries Become Democratic?” Political Science Quarterly 99 (Summer 1984), p. 303; Guillermo O'Donnell and Philippe Schmitter, Transitions from Authoritarian Rule: Tentative Conclusions About Uncertain Democracies (Baltimore and London: Johns Hopkins University Press, 1986), pp. 11; Karl, "Dilemmas of Democratization in Latin America," pp. 8-9; Frances Hagopian, "'Democracy by Undemocratic Means'?: Elites, Political Pacts, and Regime Transition in Brazil," Comparative Political Studies 23 (July 1990), pp. 147-70. 4 Guillermo O’Donnell, “Transitions to Democracy: Some Navigation Instruments,” in Robert A. Pastor, ed., Democracy in the Americas: Stopping the Pendulum (New York and London: Holmes and Meier, 1989), p. 64; Guillermo O’Donnell, “Transitions, Continuities, and Paradoxes,” in Scott Mainwaring, Guillermo O’Donnell, and J. Samuel Valenzuela, eds. Issues in Democratic Consolidation: The New South American Democracies in Comparative Perspective (Notre Dame: Notre Dame Press, 1992), p. 23; J. Samuel Valenzuela, “Democratic Consolidation in Post- Transitional Settings: Notion, Process, and Facilitating Conditions,” in Mainwaring, O’Donnell, and Valenzuela, eds., p. 76. 5 Linz, “Comment: Some Comparative Thoughts on the Transition to Democracy in Portugal and Spain,” pp. 25- 6; Karl, "Dilemmas of Democratization in Latin America"; Karl and Schmitter; Arend Lijphart, “Democratization and Constitutional Choices in Czecho-Slovakia, Hungary and Poland: 1989-91,” Journal of Theoretical Politics 4 (Number 2, 1992), pp. 207-23; Munck and Skalnik Leff; Josep Colomer, Strategic Transitions (Baltimore and London: The Johns Hopkins University Press, 2000); Omar G. Encarnación, “Labor and Pacted Democracy: Post- Franco Spain in Comparative Perspective,” Comparative Politics, 33 (April 2001), pp. 337-355. 6 O’Donnell and Schmitter; Karl, "Dilemmas of Democratization in Latin America"; Karl and Schmitter. 7 Terry Lynn Karl, "Petroleum and Political Pacts," Latin American Research Review 22 (January 1987), pp. 63-94; Frances Hagopian, "'Democracy by Undemocratic Means'?: Elites, Political Pacts, and Regime Transition in Brazil"; O’Donnell, “Transitions, Continuities, and Paradoxes”; Adam Przeworski, Democracy and the Market (New York: Cambridge University Press, 1991); Frances Hagopian, Traditional Politics and Regime Change in Brazil (New York: Cambridge University Press, 1996). 8 Theorists also hypothesize that pacted transitions depress the level of inter-party competition and reduce mass participation and inclusion. For assessments of these claims, see Bonnie N. Field, Frozen Democracy? Pacting and the Consolidation of Democracy: The Spanish and Argentine Democracies Compared (Doctoral Dissertation, University of California, Santa Barbara, 2002); Encarnación. 9 Przeworski, pp. 90-1. Przeworski distinguishes between institutional pacts and substantive pacts. However, given that the terms of the pacts are often not made public, it is difficult to assess the degree to which a pact is institutional or substantive. 10 Karl, "Petroleum and Political Pacts," p. 88.

18

11 Przeworski, p. 90. 12 Polsby uses the continuity of legislators as a measure of parliamentary institutionalization. See Nelson W. Polsby, “The Institutionalization of the U.S. House of Representatives,” American Political Science Review LXII (March 1968), pp. 144-68. 13 Share and Mainwaring; Richard Gunther, “Spain: The Very Model of the Modern Elite Settlement,” in John Higley and Richard Gunther, eds. Elites and Democratic Consolidation in Latin America and Southern Europe (Cambridge: Cambridge University Press, 1992). 14 Transition by collapse is different from revolution in that the opposition does not overthrow the incumbent regime, but rather takes control of the state by default. There is no significant violence, nor is there multilateral negotiation between the opposition and the outgoing regime. The masses are ascendant only due to the internal disintegration of the regime; they do not bring about the regime’s disintegration. 15 The analysis of the legislative branch includes all politically significant political parties whereas an analysis of the executive branch would only permit a discussion of the governing party or coalition. Also, legislators themselves are politically significant. Though actual policy may be formulated elsewhere, legislator support must normally be attained for policy to become law. 16 For an early argument that the nature of the transition in Spain led to disciplined political parties, see Jordi Capo Giol, “Estrategias para un Sistema de Partidos,” Revistas de Estudios Políticos 23 (September-October 1981), pp. 153-67. 17 The two chambers of the Argentine Congress, the Chamber of Deputies and the Senate, have relatively equal powers. However, the powers of the two chambers of the Spanish Parliament, the Cortes, are not equal. The Spanish lower house, the Congress of Deputies, is far more powerful than the Senate. The Spanish Constitution of 1978 gave the Congress the sole power to elect the prime minister and remove the prime minister from office, as well as the power to over-ride vetoes or amendments placed on legislation by the Senate. If the Senate vetoes a bill passed by the Congress, the Congress debates the Senate veto and then the initial text of the bill approved by the Congress is submitted for a vote. The Senate veto is lifted with the favorable vote of the absolute majority of the members of the Congress. If the initial text does not receive an absolute majority, another vote is taken two months from the time of the Senate veto. At this time, the veto can be lifted with a simple majority of the vote. If not, the veto stands. Senate amendments are debated and submitted to a vote in the Congress. Those that receive a simple majority of the vote are incorporated into the text. Art. 90 of the 1978 Constitution; Provisional Parliamentary Regulations for the Congress of Deputies passed October 13, 1977 (Boletín Oficial del Estado, N.256, October 26, 1977); Parliamentary Regulations for the Congress of Deputies passed February 10, 1982 (Boletín Oficial de las Cortes. Congreso, Serie H, N.33, February 24, 1982). 18 The Spanish Senate is composed of 208 directly elected senators, and a variable number of senators that are indirectly elected by the assemblies of the Autonomous Communities, currently fifty-one senators. The directly elected senators are elected from fifty-nine electoral districts, forty-seven of which are elected on the basis of a limited vote system, and the remaining districts function as plurality systems. The regional assemblies, through a system that reflects the population size of the regions and the political composition of the assemblies, elect the remaining senators. Senators are elected for a maximum four-year term (or until parliamentary elections are called). Between 1983 and 1992, Argentine senators, on the other hand, were all indirectly elected by the legislatures of the country’s 22 provinces (23 after 1990), and by an electoral college in the Federal Capital, and each elected two senators. Senators are elected for nine-year terms, with partial renewal every three years. Aside from the addition of a province after 1990, the Senate election laws were constant throughout the period studied. 19 In 1977, 207 senators were directly elected and the King appointed forty-one. Following the approval of the Constitution in 1978 and parliamentary elections in 1979, senators who were to be elected by the assemblies of the Autonomous Communities were incorporated as decentralization and regional elections occurred. Full incorporation of these representatives did not occur until midway through the 1982/86 legislature. 20 The minimum threshold in Spain is 3% of the valid votes, while the Argentine minimum threshold is 3% of registered voters. 21 Ceuta and Melilla, Spanish territories in northern Africa, elect only one deputy each. 22 Ceuta and Melilla are not considered provinces.

19

23 Ley para Reforma Política 1/1977, Real Decreto-Ley Sobre Normas Electorales 20/1977, Art. 68-9 of the 1978 Constitution, and the Régimen Electoral General Ley Orgánica 5/1985. The latter law was later modified by Leyes Orgánicas 1/1987 and 8/1991. 24 254 deputies were elected in 1983, half of which served two-year terms and the other half of which served full four-year terms. 25 The Argentine data include deputies elect that held a seat in the previous legislative period regardless of whether they were in office at the close of the prior legislative term. This only affects the calculations in two elections, 1989 and 1991. The inclusion of only those deputies that were sitting at the close of the prior legislative term slightly reduces the level of reelection. 26 During the same period, the average percentage of Spanish senators reelected was 40 percent. In Argentina, the average was 35 percent in the Senate. Sources: The Spanish average was calculated using 1977-1986 reelection data published in María Luz Morán, “Un Intento de Análisis de la ‘Clase Parlamentaria’ Española: Elementos de Renovación y de Permanencia (1977-1986),” REIS 45 (January-March 1989), pp. 61-84. The 1989 Spanish reelection rate was calculated by the author based on data provided by Mariano Baena. The Argentine rate was elaborated by the author based on senator lists provided by the Argentine Congress. 27 This election did not coincide with a presidential election, and confirmed the relative popularity of the governing Unión Cívica Radical. 28 The Pedersen index “measures the net change in the seat (or vote) shares of all parties from one election to the next. The index is derived by adding the net change in percentage of seats (or votes) gained or lost by each party from one election to the next, then dividing by two. An index of 15, for example, means that some parties experienced an aggregate gain of 15 percent of the seats from one election to the next while others lost a total of 15 percent.” Scott Mainwaring and Timothy R. Scully, “Introduction: Party Systems in Latin America,” in Scott Mainwaring and Timothy R. Scully, eds. Building Democratic Institutions (Stanford: Stanford University Press, 1995), p. 6. 29 Spanish volatility indexes were taken from Eva Anduiza and Mónica Méndez, “Elecciones y Comportamiento Electoral,” in Manuel Alcántara and Antonia Martínez, eds., Política y Gobierno en España (Valencia: Tirant lo Blanch, 1997). Kenneth M. Roberts and Erik Wibbels, University of New Mexico, provided the Argentine volatility indexes. Also see Kenneth M. Roberts and Erik Wibbels, “Party Systems and Electoral Volatility in Latin America: A Test of Economic, Institutional, and Structural Explanations,” American Political Science Review 93 (September 1999), pp. 575-90. 30 Excluding the 1982 outlier, the Spanish average declines to eight, which puts the level of volatility below the Argentine average. However, the Spanish elections in 1986 had the same degree of volatility as the 1987 Argentine elections, and 57 percent of the Spanish deputies were reelected versus only 24 percent of the Argentine deputies. Furthermore, the extremely volatile 1982 Spanish elections (43 volatility) produced a reelection rate of 40 percent, a level of reelection never reached in Argentina despite significantly lower levels of volatility. 31 1993 data reelection rate was calculated by the author based on deputy lists provided by the Congreso de los Diputados, Spain. 32 These figures include deputies prior to the 1989 elections in Spain, and prior to the 1995 elections in Argentina. Terms may be partial as the analysis includes all deputies elect and substitutes. The Spanish data include 1400 deputies elect and 139 substitutes. The Argentine data include 892 deputies elect and 119 substitutes. 33 In Argentina, deputies holding seats during the entire period could have been deputies during four terms (1983/85, 1985/89, 1989/93, 1993/97) or three (1983/87, 1987/91, 1991/95). 34 Scott Mainwaring and Matthew Soberg Shugart, “Conclusion: Presidentialism and the Party System,” in Scott Mainwaring and Matthew Soberg Shugart, eds. Presidentialism and Democracy in Latin America (Cambridge and New York: Cambridge University Press, 1997), pp. 421-29. 35 Richard S. Katz, “The Problem of Candidate Selection and Models of Party Democracy,” Party Politics 7 (N.3, 2001), p. 277. 36 The election results are based on data provided by the Junta Electoral Central, Ministerio del Interior, Spain. 37 N. Guillermo Molinelli, M. Valeria Palanza and Gisela Sin, Congreso, Presidencia y Justicia en Argentina: Materiales para su Estudio (Buenos Aires: CEDI –Fundación Gobierno y Sociedad, 1999), pp. 268-75. 38 For additional discussion of party rules in Argentina, see Mark P. Jones, “Evaluating Argentina’s Presidential Democracy: 1983-1995,” in Scott Mainwaring and Matthew Soberg Shugart, eds., Presidentialism and Democracy in Latin America (New York: Cambridge University Press, 1997); Miguel De Luca, Mark P. Jones, and María Inés 20

Tula, “Argentine Political Parties and their Candidate Nomination Procedures,” paper prepared for the XXII International Congress of the Latin American Studies Association, Miami, Florida, March 16-18, 2000; Mariano Tommasi and Pablo T. Spiller, Las Fuentes Institucionales del Desarrollo Argentino. Hacia una Agenda Institucional (Buenos Aires: Editorial Universidad de Buenos Aires, 2000); Miguel De Luca, Mark P. Jones, and María Inés Tula, “Back Rooms or Ballot Boxes? Candidate Nomination in Argentina,” Comparative Political Studies 35 (May 2002), pp. 413-36; Mark P. Jones, “Explaining the High Level of Party Discipline in the Argentine Congress,” in Scott Morgenstern and Benito Nacif, eds., Legislative Politics in Latin America (Cambridge and New York: Cambridge University Press, 2002); Mark P. Jones, Sebastian Saiegh, Pablo T. Spiller, and Mariano Tommasi, “Amateur Legislators-Professional Politicians: The Consequences of Party-centered Electoral Rules in a Federal System,” American Journal of Political Science 46 (July 2002), pp.656-69. 39 Jones, “Explaining the High Level of Party Discipline in the Argentine Congress”; Jones, Saiegh, Spiller and Tommasi. 40 Jones, “Explaining the High Level of Party Discipline in the Argentine Congress, ” pp. 159-62. 41 Jones, Saiegh, Spiller and Tommasi, p. 658. 42 This may be due to disagreement between the provincial and national party organizations or because the provincial party organization cannot agree on a single list. Jones, “Evaluating Argentina’s Presidential Democracy: 1983-1995,” pp. 270-75. 43 Unión Cívica Radical, “Carta Orgánica Nacional,” (1985, 1989, 1993 and 1997), Art. 31. The list on which the candidate appears must receive 2/3 of the vote in the primary. If the list does not receive the required 2/3, the next person on the list replaces the sitting deputy. I thank Mark Jones for this clarification. 44 De Luca, Jones, and Tula, “Back Rooms or Ballot Boxes? Candidate Nomination in Argentina.” 45 Ricardo Gutiérrez, “Desindicalización y Cambio Organizativo del Peronismo Argentino, 1982-1995,” paper prepared for presentation at the XXI International Congress of the Latin American Studies Association, Chicago, September 1998. 46 All data in this paragraph are taken from De Luca, Jones, and Tula, “Back Rooms or Ballot Boxes? Candidate Nomination in Argentina.” 47 De Luca, Jones, and Tula, “Argentine Political Parties and their Candidate Nomination Procedures,” p. 21. 48 Unión Cívica Radical, “Carta Orgánica Nacional,” (1985, 1989, 1993 and 1997), Art. 31. Partido Justicialista, “Carta Orgánica Nacional. Aprobada según las modificaciones introducidas en el Congreso Nacional Partidario, realizado el 20 de Septiembre de 1991,” Art. 44. 49 De Luca , Jones and Tula, “Back Rooms or Ballot Boxes? Candidate Nomination in Argentina,” p. 420. 50 According to Mark Jones, this occurs in approximately 2/3 of the cases. Personal communication. 51 In 1998, the PSOE decided to use primaries to select some candidates, particularly the heads of the candidate lists, as part of the party’s strategy to renew itself. (Partido Socialista Obrero Español, “Regulations for the Selection of Candidates for Public Office,” approved by the PSOE Federal Committee in March 1998, modified in July 1998, and used for the 2000 elections.) The primaries slipped out of control when put into practice to determine the party candidate for Prime Minister. Josep Borrell defeated the party hierarchy candidate, Joaquín Almunia. Borrell later stepped down and was replaced by the loser of the party primary - Almunia. Almunia, in the parliamentary elections held on March 12, 2000, led the party to its worst electoral result since 1979. 52 This process further reduces the participation of rank-and-file party members, who are most likely to influence outcomes at the local level. 53 Partido Socialista Obrero Español, “Estatutos del Partido Socialista Obrero Español,” passed in the 28th through the 34th Party Conventions between May 1979 and June 1997. PSOE internal party documents: “Procedure for the Elaboration of Lists of Candidates for Legislative Elections” used for the 1986, 1989, 1993, and 1996 elections. Unión de Centro Democrático, “Estatutos de Unión de Centro Democrático,” passed in the first (October 1978) and second (February 1981) party Congresses. The candidate selection procedures in Alianza Popular/Partido Popular were very similar to those of the UCD and PSOE. Between 1979 and 1996, regional or provincial parties proposed candidate lists, but a national-level party body had the ability to alter the lists and make the final determinations. Elena María García-Guereta Rodríguez, Factores Externos e Internos en la Transformación de los Partidos Políticos: el Caso de AP-PP (Doctoral Dissertation, Centro de Estudios Avanzados en Ciencias Sociales, Madrid, 2001), p. 486. 54 Richard Gillespie, The Spanish Socialist Party (New York: Oxford University Press, 1989), p. 343.

21

55 The Federal Committee is composed of members of regional party federations and members of the Executive Committee. 56 Mónica Méndez Lago, La Estrategia Organizativa del Partido Socialista Obrero Español (1975-1996) (Madrid: Centro de Investigaciones Sociológicas, 2000), p. 124. My own translation. 57 Art. 24. Reproduced in Manuel Contreras, El PSOE en la II República: Organización e Ideologia (Madrid: Centro de Investigaciones Sociológicas, 1981), p. 295. 58 This led to a great deal of continuity between authoritarian political class and UCD politicians under the new democracy. Field, pp. 80-112. 59 Jonathan Hopkin, Party Formation and Democratic Transition in Spain: The Creation and Collapse of the Union of the Democratic Centre (New York: St. Martin’s Press, 1999), p. 56. 60 Richard Gunther, Giacomo Sani, and Goldie Shabad, Spain After Franco: The Making of a Competitive Party System (Berkeley: University of California Press, 1988), p. 100. 61 Unión de Centro Democrático, “Estatutos de Unión de Centro Democrático,” passed in the first (October 1978) party Congress, Art. 45.2.C. and Art. 21.1. 62 Hopkin, p. 89. 63 Ibid., p. 96. 64 Ibid., p. 97. 65 Unión de Centro Democrático, “Estatutos de Unión de Centro Democrático,” passed at the second (February 1981) party Congress. My own translation. 66 While the PSOE and UCD candidate selection procedures permit national elite control over candidate lists, it is difficult to systematically measure the degree to which national political elites intervened in the selection of candidates. Any measure would have to take into account formal intervention as well as the effect of potential intervention on the selection of candidates and their position on the list. 67 Gillespie, The Spanish Socialist Party, p. 346. 68 Partido Socialista Obrero Español, “Estatutos del Partido Socialista Obrero Español,” 27th (1976) through 34th Party Convention (1997), Art. 3. 69 Richard Gillespie, “Factionalism, the Left and the Transition to Democracy in Spain,” in Richard Gillespie, Michael Waller and Lourdes López Nieto, eds., Factional Politics and Democratization (London: Frank Cass, 1995), p. 47. 70 In 1984, however, the party explicitly began to permit, regulate, and provide for the representation of ‘currents of opinion’ within the party. (Partido Socialista Obrero Español, “Estatutos del Partido Socialista Obrero Español,” 30th Party Convention, 1984, Art. 3.) Most importantly for our analysis, the party regulations provided for their representation on non-executive bodies, such as Convention Delegations, and Provincial, Regional and Federal Committees. These bodies would be elected in a way that facilitated the representation of these ‘currents.’ A minority that obtained at least 20 percent of the votes would attain 25 percent of the positions up for election, and, if two minorities achieved 20 percent of the vote, each would obtain 20 percent of the positions. However, this did not include lists for public office, nor did it include executive bodies, which continued to be elected using a majoritarian system. 71 Unión de Centro Democrático, “Estatutos de Unión de Centro Democrático,” passed at the first (October 1978) and second (February 1981) party Congresses. 72 Hopkin, pp. 91, 138. 73 For each electoral district list, the candidates are presented by political party or electoral coalition. Each political party presents their candidates and alternates in the order in which the seats will be distributed. This measure includes those candidates on the list that equal the number of seats available in that district, i.e. alternates are not included. For the parties analyzed, the data include the party-list itself or any list presented by a formal electoral coalition in which the party under study took part. Independents that are included on the party-list are also included in the analysis. 74 The two largest parties were determined by looking at the two largest parties (% of seats) in the previous election, i.e. the 1982 parliamentary election average in Spain included the two largest parties from the 1979 parliamentary elections. If the analysis were to include the two largest parties resulting from the election under scrutiny, the average candidate continuity in Spain would decline slightly to 29 percent. 75 There is not a directly comparable measure of candidate continuity in both cases. All Spanish deputies are elected when parliamentary elections are called. However, Argentine deputies are elected for a four-year term, and 22 half of the Chamber is renewed every two years (except in 1983 when the entire Chamber was elected, half of the Chamber served for two years and the other half for four years). This means, for example, that if a candidate is not elected on the list presented in 1985, he can be placed on the list again in 1987, and again in 1989. Therefore, we cannot simply compare candidate lists that are four years apart because we would miss those that reappear in between. I have, therefore, compared the proportion of Spanish candidates on the list under scrutiny that were also candidates in the previous election to the proportion of Argentine candidates that were candidates in either of the previous two elections. Though this measure may slightly boost Argentine deputy continuity, it represents a more difficult test of whether pacting produces democracies in which there is a greater degree of elite continuity. 76 All Spanish candidate re-nomination data were calculated by the author based on candidate lists published in the Boletín Oficial del Estado. June 15, 1977 elections: Candidate lists (May 20, 1977) and corrections (May 23, 24, 25, 26, 27, 28, 30, 31, and June 1, 2, 3, 4, 6, 7, 8, 9, 11, 14, 15, 17, 18, 20). March 1, 1979 elections: Candidate lists and corrections were reproduced in Jorge de Esteban and Luis López Guerra, eds., Las Elecciones Legislativas del 1 de Marzo de 1979 (Madrid: Centro de Investigaciones Sociológicas, 1979). October 28, 1982 elections: Candidate lists (October 2, 1982) and corrections (October 9, 14, 16, 27, 28). June 22, 1986 elections: Candidate lists (May 21, 1986) and corrections (May 27, 28, 29, 30 and June 3, 6, 14, 20, 21). October 29, 1989 elections: Candidate lists (September 30, 1989) and corrections (October 7, 14, 21, 28). All Argentine candidate re-nomination data were elaborated by the author based on candidate lists collected from the Departamento de Estadísticas, Dirección Nacional Electoral, Ministerio del Interior, Argentina. 77 Jones, “Explaining the High Level of Party Discipline in the Argentine Congress,” pp. 177-78. 78 Capo Giol; Mario Caciagli, “Spain: Parties and the Party System in the Transition,” in Geoffrey Pridham, ed., The New Mediterranean Democracies: Regime Transition in Spain, Greece and Portugal (London: Frank Cass and Company Limited, 1984). 79 Alfred Stepan, “Paths Toward Redemocratization: Theoretical and Comparative Considerations,” in Guillermo O’Donnell, Philippe C. Schmitter, and Laurence Whitehead, eds., Transitions from Authoritarian Rule: Comparative Perspectives (Baltimore and London: The Johns Hopkins University Press, 1986), p. 80. 80 Field, 2002. 81 Capo Giol, p. 159. 82 For a discussion of party discipline in the Congress of Deputies, see Manuel Sánchez de Dios, “La Disciplina de Partido en los Grupos Parlamentarios del Congreso de Diputados,” Revista de las Cortes Generales 39 (Third Quarter 1996). 83 Capo Giol, p. 159. 84 Caciagli, “Spain: Parties and the Party System in the Transition,” p. 85. 85 Members in the interior of Spain. Jorge de Esteban and Luis López Guerra, Los Partidos Políticos en la España Actual (Barcelona: Editorial Planeta, 1982), p. 116. 86 Méndez Lago, p. 56. 87 Guerra was first named Secretary of Party Organization and later Vice-Secretary General. 88 Esteban and López Guerra, Los Partidos Políticos en la España Actual, p. 117. 89 Paul Heywood, “Mirror-images: The PCE and PSOE in the Transition to Democracy in Spain,” West European Politics 10 (April 1987), p. 206. 90 Mario Caciagli, Elecciones y Partidos en la Transición Española (Madrid: Centro de Investigaciones Sociológicas, 1986), p. 209. 91 Gillespie, The Spanish Socialist Party, p. 338. 92 Ibid., p. 395. 93 Ibid, p. 343, 401; Caciagli, Elecciones y Partidos en la Transición Española, p. 211. 94 Méndez Lago, p. 112. 95 Ibid., p. 127. 96 Heywood, pp. 206-7. 97 This Partido Popular should not be confused with the Partido Popular founded in 1989. The latter was the name adopted by Alianza Popular due to its re-founding. 98 Esteban and López Guerra, Los Partidos Políticos en la España Actual, p. 91. 99 Hopkin, p. 53. 100 Caciagli, Elecciones y Partidos en la Transición Española, p. 252. 101 Esteban and López Guerra, Los Partidos Políticos en la España Actual, p. 93. 23

102 Ibid., p. 92. 103 Emilio Attard, Vida y Muerte de UCD (Barcelona: Planeta, 1983), p. 56. 104 Beate Kohler, Political Forces in Spain, Greece and Portugal (London: Butterworth, 1982), p. 32. 105 Hopkin, p. 86. 106 Esteban and López Guerra, Los Partidos Políticos en la España Actual, p. 94. 107 Ibid., p. 95. 108 Ibid. 109 Ibid., p. 93. My own translation. 110 Caciagli, “Spain: Parties and the Party System in the Transition,” p. 85. 111 De Luca, Jones, and Tula, “Back Rooms or Ballot Boxes? Candidate Nomination in Argentina,” p. 425. 112 Caciagli, “Spain: Parties and the Party System in the Transition,” p. 89. 113 Przeworski, p. 90. 114 Field, pp. 196-243.

24